Eating Value

dont get me wrong, i have nothing against getting value, but it does annoy me when people i know say they fancy a horse, i see them later and say, i had a nice win on that horse and they say, i didnt back it cos i wanted 3/1 and it was only 2/1, ok, so i get the drinks in shall i.

let's say, hypothetically, federer plays murray tomorrow. you think federer has a 70% chance of winning. the odds are 1.30 federer, 4.33 murray. sophie and schalke, who do you back

tobermory 29 Mar 01:10

But who is it that 'can't get a winner' ?

everyone gets winners , unless they only bet once a year on the national .

You give an example of just one race , but there are thousands of races every year to pick from , if someone is backing 3/1s that should be 5s , over the year they will have the same number of wins as the guy backing 10/1 that should be favs . Easy to see who will have the most money at the end of it.

Lots of people fail at gambling because they don't pick enough winning bets .... thats how bookies make their money !! But if someone comes along and constantly backs winners then they are the opposite ....

tobermory 29 Mar 01:10


But who is it that 'can't get a winner' ?

everyone gets winners , unless they only bet once a year on the national .

You give an example of just one race , but there are thousands of races every year to pick from , if someone is backing 3/1s that should be 5s , over the year they will have the same number of wins as the guy backing 10/1 that should be favs

why will they have the same number of winners

A treeeemendous machiiine 29 Mar 01:12

let's say, hypothetically, federer plays murray tomorrow. you think federer has a 70% chance of winning. the odds are 1.30 federer, 4.33 murray. sophie and schalke, who do you back

Federerof course, Murray admitted today that he has "gone" mentally ... in terrible form.

It is not about number of overall winners though , that is irrelevant , only the number of winners compared to the average prce they back at .

if both the 3/1s and the 10/1s should be 5/1 then they will both win 1 in 6

A treeeemendous machiiine 29 Mar 01:12


let's say, hypothetically, federer plays murray tomorrow. you think federer has a 70% chance of winning. the odds are 1.30 federer, 4.33 murray. sophie and schalke, who do you back

i back who i think will win,

what if federer was 1.05 and murray was 21. still federer?

fair enough sophie, we'll both keep trying to pick the winners :D

I wouldn't take Federer at 1/05, but I wouldn't take Murray at 21's either, cause I don't think he will win ... but the odds on Federer have put me off the bet

tobermory 29 Mar 01:15


It is not about number of overall winners though , that is irrelevant , only the number of winners compared to the average prce they back at .

not at all, why should they both win the same amount of times just because he got value.
i might have 20 sh it bets that win, he might have 3 value ones that win,
i think if you get a winner its value and the bigger price you get its more value.

But "if I had to" I would take Federer at 1.05, 1.01 if need be .... I think he will win no matter what ! But yeah, of course the price comes into it, but no value in Murray whats so ever if I don't think he will win eh !!

good, not taking federer at 1/20 is a good start :) i'll never be able to convince you to back underdogs, but each to their own

i cant stand the miserable scottish git,

sophiep

You are like a breath of fresh air to the forum !! Finally there is someone that see's it as it is unlike 98% of the morons on here that think it's all about value lol

A treeeemendous machiiine 29 Mar 01:21


good, not taking federer at 1/20 is a good start i'll never be able to convince you to back underdogs, but each to their own

I will back an underdog if I think they will win, but, you have to remember that bookies very rarely get it worong !! They price favs for a reason, and bookies doing that actually is a good way of picking winning bets .... odds on win more that longer odds, don't they !! If you start picking constant longer odd bets then it's that little bit tougher to get the winning ones, thats what I have found in my experience at betting ... good value odds on winning bets is what I like.

If you are winning overall backing all those 3/1s Sophie , then the guy who was saying they should be 5/1 got it wrong , and you got the value.

Is CF AMERICA odds at 12.5 in Match market "value" ??????Currently 1-0 PUEBLA.....

I think it is!

Go back it then

Fixtures 28 March / Puebla v CF America / Match Odds
Puebla Lay29-Mar-10
01:26 1.53 36.00 1.53 29-Mar-10 01:27

Fixtures 28 March / Puebla v CF America / Match Odds
CF America Back 29-Mar-10
01:26 14 6.00 14 29-Mar-10 01:26

Or is backing PUEBLA??????

What is your opinion SCHALKE?

Yes Schalke , more 1/2s win than 2/1s , so if you just back the favs you get more wins , but if 63% of the 1/2s and 36% of the 2/1s win, then the 2/1 backer is ahead , despite havin much fewer winning bets!

Don't know either of them, don't know the time of match so I'm not touching it and unable to give an opinion, sorry

Fair enough ........it is on BF VIDEO now...and CF A have missed umpteen chances to score ....but they are still creating chances ...and are likely to score 1 or 2......IMO THEY ARE VALUE!...AS IS THE DRAW

I will back an underdog if I think they will win, but, you have to remember that bookies very rarely get it worong !! They price favs for a reason, and bookies doing that actually is a good way of picking winning bets .... odds on win more that longer odds, don't they !! If you start picking constant longer odd bets then it's that little bit tougher to get the winning ones, thats what I have found in my experience at betting ... good value odds on winning bets is what I like.

as an odds compiler, i guess i should take that as a compliment :P i can assure you, i can get it wrong, and sid james and bet £3.65 to name a couple get it horrendously wrong on a regular basis. I'm off to bed, have a good night :D

Yeah, totally agree and thats where your opinion on the sports you know are vital to get the bets that they price wrong ... you still got to hope for a winner though,

Bookies have huge overounds , so they have a big margin for error in their pricing. They usually still win when they are wrong.

2-0 PUEBLA ....but there was also value in 2-0 ]:) :D

not when you are as wrong as sid

Thats where I like looking at bookie prices then backing on here :D

DRAT 2-1 ...and I never got my lay on 2-0 taken DRATTT :(

Still have my 2 value bets on though :D

Keeper just prevented equaliser ...so whether I win or lose ....my bets were ...ARE ..."value"!

I did back 2-0 as a cover at 1-0 .....

Fixtures 28 March / Puebla v CF America / Correct Score
2 - 0 Back 29-Mar-10
01:32 7.2 4.00 7.2 29-Mar-10 01:32

So the 3 bets were value ...................but me being too slow on lay on 2-0 was just pure "bad luck" due to speed of 3rd goal going in to make it 2-1.


You might argue 1.54 for PUEBLA was value at 1-0...............but not if you were watching game !

G M the noooo

Have all the poeple here who think value is a nonsence concept made a killing from gambling? I assume so, being as they can predict the outcome of events, the price of which is largely irrelevant in their eyes. They must be raking it in :)

I doubt they'll post up 3 months P&L, but they'll know if it's overall negative, and I wonder what would happen to that figure if they look back over the bets and add 10% to the winning backs and deduct 10% from the losing lays? If that turns the negative figure into a positive figure, could they achieve that positive figure in the future by getting a better value price when placing their bets?

Bayes. 27 Mar 12:19

Let's say I get 2.1 on tails for 10,000 individual tosses of a coin. I win exactly half and show a nice profit. Clearly I have obtained value on each and every bet regardless of outcome.

Now let's say I am a good judge of sporting events. I find 10,000 events where I can obtain 2.1 about what I subjectively make a 2.0 shot. Again I win exactly half and show a nice profit. The question is, did I obtain value on each and every bet, or did I obtain more value on some and less on others?

I would say that from the information given above it's impossible to say. Let's say that on each of the 10,000 events you bet £2, meaning that you either lose £2 or win £2.20 each event (ignoring commission). After 10,000 events you could be up £600, you could be up £1,400, and you could be losing (minuscule chance, but if you changed the odds to 2.05 this chance becomes far higher).

In the first instance it's easy to prove (practically) whether you are getting value (and exactly how much) by determining that the coin is fair. In the second instance this is not a possibility as the events are too complex. You can establish that you have an edge with a high degree of certainty, but objectively speaking you cant' be absolutely sure, although you can get to a stage that for all practical purposes you can view it as fact.

You may be able to show that the chance of your results being due to luck is lower than 0.01% or some other percentage by analyzing performance metrics, but it can never be zero (again for all practical purposes you could view it as such, in the same way that you don't go around analyzing and avoiding billion to one risks of death)

Betting is a combination of skill and luck/randomness. Anyone that doesn't believe this should look at the above again. If you have two people who are able to back 2.0 shots at 2.1 over a million bets with the exact same staking (obviously the two sets of one million bets are separate) their terminal wealth could be very substantially different. If a million bets isn't long term I don't know what is. Anyone that states 'in the long run...' is in danger of forgetting that 'the long run' may be so long that the expected results do not necessarily occur during a lifetime.

Again all this could be stated with confidence intervals, but not with certainty. Even though you can state with something close to certainty that someone who has a large edge will win over an extended series of bets, terminal wealth could be x, 2x, 3x etc. based on luck. The more value you have, the less important the effect of luck/randomness becomes.

The skill part equates to finding value. The fact that it's possible (though unlikely) to win long term without value is what leads people to the belief that value doesn't matter. It's possible for someone to win the lottery. It can be shown that it is equally probable for someone to come out with a large profit on a long series of small bets due to 'dumb luck'. For instance, if 1000 people were to carry out the series of 10,000 bets as above, their results would be spread out, some would make £500, some would make £1500 etc. If the odds obtained were higher, the differences would be expected to diminish.

Equating value to opinion does not seem natural to me. Opinion is more rigid, any idea of value changes constantly as new information becomes available. Value is about making money, opinion is about being right.

Schalke 04 29 Mar 01:22
sophiep

You are like a breath of fresh air to the forum !! Finally there is someone that see's it as it is unlike 98% of the morons on here that think it's all about value lol




This is an illustration of why it is so easy to make money here. Bless his cotton socks.

My plan is to help Schalke and all his followers to throw out of the window the value concept. It is quite clearly only about picking winners so this one is sorted innit.

Essentially everyone on the schalke side of the argument is in the same boat:

They think value is nothing because they don't approach betting from the standpoint that the people on the other side of the debate do. They try to pick winners, and essentially haven't got a clue whether the price is correct or not. They have bets at short prices which they would have had at 5, 10, 20 ticks lower and don't really have a definite cutoff point for a bet. I'm not saying they will go round and back things at 1.01 because one has to assume they have SOME kind of cutoff point, but they don't actually identify what it is before they place it, it would be more of a knee-jerk feel reaction.

As identified earlier in the thread, that doesn't make it impossible that they will win. The more bets they have, the more unlikely it becomes that they will finish on top, IMHO.

Surprised betfairy is quiet because he is, if the forum is to be believed (sic) a big winner who was scant regard for attempting to price things correctly.

The other side of the debate tend to approach their betting from a point of view of having an event priced up (however they go about doing it) and betting when the odds are out of line with their own figures as a general rule. They might do it once per match/race and move on, they might back/lay multiple runners, they might make many trades in running, or even do other things.

The fact of the matter is, like it or not, that the first pool of people contains net losers and the second pool of people contains net winners. Its just whether you accept that as a coincidence, or not. The first pool of people interestingly often contains people who have ATTEMPTED to make it into the second pool, but it hasn't worked out for them, and that's generally because they can't produce accurate prices for an event, or they overstake, or fall into one of the many other pitfalls.

Schalke 04 29 Mar 01:13

Lots of people fail at gambling because they don't pick enough winning bets .... thats how bookies make their money !! But if someone comes along and constantly backs winners then they are the opposite ....

So why are you down to your last £40? Where did it all go wrong?

Spring is late this year but the cherry blossom and the daffs are out now. And along with Spring comes the first thread on "value" since...well, the last thread on value. It's not been seen for a while but its a hardy annual and arrives every year in time for the Flat Turf season. It usually thrives throughout the summer but many people do not believe in it at all. They think its like the toothfairy or Father Christmas. Others believe in it passionately and claim that those who don't believe are blind to the obvious.


Belief in "value" does not seem to be a function of intelligence. I've know many a university educated person who can't see it whilst I've also know clever people who believe that they can predict roulette numbers. It takes all sorts.

Remember, it wasn't so long ago that people thought that the Earth was flat or that kings were appointed by God. Indeed, many still believe in a God. Who is to say that they are wrong?

For those of us (and I include myself in this this group) who believe that recognising the importance of the potential profit contained within a play (value) is key to l/t success, remember that keeping an eye on the speedometer whilst driving is important, but there is a danger in that too much focus on the speed will blind you to whats obvious in front of you on the road.

And to those who don't wish to use/understand value, but seem to drive successfully, I would say that you risk going over the speed limit because its not always possible to judge your speed by the eye alone.

Happy motoring one and all.

Who opened this can of worms? Can we have a discussion on the merits of weight affecting a horses performance next? That one's always good for a laugh.

remember that keeping an eye on the speedometer whilst driving is important

A racing car doesn't have a speedo.

My take on the original question is that very likely there will be some bets where you had more value than others.

However, if you were to continue betting £2 on another 10,000 events where you were confident (given your previous results) that you could get the same average value, then it would make no difference to your "risk management" calculations and you could happily assume that each bet independently offered the same value.

The problem will come when you want to increase stakes, try getting £10 on each of the 10,000 events and you will likely find that the ones that are harder to fill your order on, were the ones with the best value, thus your average value per bet will be reduced and if you increase your stakes too much, the average value per bet could even go into negative territory.

That then starts the next round of analysis on your pricing etc. etc.

zipper 27 Mar 12:32


The only value is the Winner .. price dont matter ... trust me
=======================================================


Selection Odds Stake(£) Bid type Placed Profit/loss(£)
Anquetta 1.13 24.00 Lay 26-Mar-10 12:54 24.00
Ceepeegee 3.50 3.00 Back 26-Mar-10 12:52 7.50
Dont Tell Nina 6.84 4.00 Back 26-Mar-10 12:51 23.35
Pendower 1.27 24.00 Lay 26-Mar-10 12:55 -6.48
*Average odds: On Off Back subtotal: 30.85
Lay subtotal: 17.52
Market subtotal: 48.37
Commission @ 3.49%: 1.69
Net Market Total: 46.68
=====================================================

Anquetta 1.77 3.00 Back 26-Mar-10 12:52 -3.00
Anquetta * 1.78 26.00 Lay 26.00
Ceepeegee 38.00 2.00 Back 26-Mar-10 12:52 -2.00
Ceepeegee 6.00 12.00 Lay 26-Mar-10 12:53 12.00
Dont Tell Nina 100.00 3.00 Back 26-Mar-10 12:50 -3.00
Dont Tell Nina 6.00 33.00 Lay 26-Mar-10 12:53 33.00 *Average odds: On Off Back subtotal: -8.00
Lay subtotal: 71.00
Market subtotal: 63.00
Commission @ 3.51%: 2.21
Net Market Total: 60.79
======================================================




RESULT
1 Pendower (IRE) 11-1 R Walsh P F Nicholls 6 7/2
raced on outside, led approaching 2nd, headed narrowly, still travelling well when not fluent 3 out, led again 2 out, ridden clear flat, ran on well opened 9/4 touched 7/2
2 7 Ceepeegee (IRE) 11-1 J Tizzard C L Tizzard 5 16/1
set slow pace until before 2nd, remained close up, led approaching 2 out, soon headed, 1 length down and ridden when mistake last, no impression opened 25/1 touched 25/1
3 3¼ Don't Tell Nina (IRE) 11-1 W Kennedy P R Webber 6 33/1
held up in rear, headway 4th, led before 3 out, headed approaching 2 out, 3rd and beaten last, one pace opened 50/1 touched 50/1
4 6 Supreme Plan (IRE) 11-1 S Quinlan (3) K C Bailey 7 16/1
tracked leaders, mistake 6th, every chance when not fluent 3 out, hung right and weakened approaching last opened 20/1 touched 20/1
5 6 Anquetta (IRE) 11-1 B J Geraghty N J Henderson 6 1/2 f held up towards rear, headway when jumped left and not fluent 7th, shaken up after 3 out, went left and beaten 2 out, soon weakened, went left last opened 4/6 touched 4/6 £800-£1100 £200-£300 (x2) £400-£700 £427-£800 £200-£400

6 2¾ Grand Award 11-1 R Thornton D E Cantillon 5 12/1
mid-division, headway approaching 4th, steadied next, progress before 3 out, switched left approaching 2 out, soon hanging left, weakened between last 2 opened 14/1 touched 16/1
7 12 Robin De Paques (FR) 11-1 Mr D H Dunsdon (3) N J Gifford 6 100/1
tracked leaders, lost place before 6th, pushed along next, weakened approaching 3 out opened 100/1 touched 100/1
8 21 Honourable Dreamer (IRE) 11-1 A P McCoy Jonjo O'Neill 5 33/1
prominent until not fluent 3rd, not fluent and dropped to rear 5th, weakened approaching 3 out, tailed off opened 50/1 touched 50/1
9 30 Frankly Fantastic 11-1 t Felix De Giles A J Lidderdale 6 100/1
in rear, not fluent 4th, weakened approaching 3 out, tailed off opened 100/1 touched 100/1

Non-Runners
NR Foxhill 11-1 N Runner Mrs J G Retter 10


Tote Win: £3.60 Tote Place: £1.40,£2.30,£7.90
Straight Forecast: £49.84 Exacta: £34.50
Swingers:1&2:£6.90, 2&3:£28.80, 1&3:£16.00
=======================================================


So basically zipper -

1. You don't have to get the winner to obtain "value".

EG the 3rd horse was layed IR to create a profit in WIN Market.

EG the 2nd and 3rd horses can be backed to PLACE , at value odds" AND YOU "WIN".

EG the favourite did not "WIN" and was layed .....so value was obtained on laying it at very very low value odds NOT TO WIN.

2. Price does "matter", because if the odds of horses that came in 2nd and 3rd were too low in "value" .......say odds on in place market.......then I just would not have "risked" my money on a BACK bet to place.

I can give you MANY MANY examples where you DON'T HAVE TO GET THE WINNER TO WIN.........but they are almost always on horses at VALUE odds ( eg 40-1 SP upwards in place market)

galejo,

I wondered when someone would answer Bayes' question, cheers :)

I think he's experimenting on the forum to see how quickly it degenerated into a discussion about Schalke04 and co.

Good point about being harder to get filled in on the lesser bets and the continual fight to re-analyse stuff.

Coin toss, coin toss, coin toss....yawn

The other day, I laid Arsenal away to Birmingham, I fancied Brum might get something from the game as they have done against all the other top sides at St Andrews this season. Did I get value? I don't know, I don't think I'm qualified to judge. I was just happy with the potential risk/reward of the bet, but I wouldn't say 'the market' priced it up incorrectly.

That Arsenal side have played that Birmingham side once at St Andrews this season, and that is the only time that will ever happen, under the same conditions, with the same fixture distractions (Arsenal with Barca mid-week) etc etc. I don't get the benefit of this event happening 10 thousand times over.

A coin-toss analogy is a f*cking nonsense, because a coin toss can be done indentically thousands of times over. It's just a facile analogy, when it comes to Sports Betting.

BTW - you can back a horse at VALUE odds to win ....and it comes 2nd.....and you might say "you can't eat value"...........BUT YOU CAN WHEN YOU ALSO BACK IT IN SFC/EXACTA TO BE 2ND TO THE HORSE THAT IS LIKELY TO BEAT IT!

If you did this with low odds ...poor value horse ......you would LOSE over all , as when you get a SFC/EXACTA the returns would not cover your outlays.

There's a lot of tosh spoken about value ... Particularly by some of the idiots on Channel 4 racing ... Its a relative term, a price one person may see as 'value' another may not ... and if not of your 'value' bets win,none of your bets are worth squat

98% lose............................why? They tend to back low value odds horses.........

Schalke 04 = complete an utter mug wagon!

End of

Equating value to opinion does not seem natural to me. Opinion is more rigid, any idea of value changes constantly as new information becomes available. Value is about making money, opinion is about being right.

To show what I mean by this, imagine that someone posted on the forum that a 1.1 shot is good value (and should be priced at 1.08). If this bet goes on to lose, a lot of people would be eager to point out what a mug this person is. If it wins, sarcastic comments like 'well done backing at 1.1 and winning' would come up. If the bet loses people will see this as proof that your 'opinion' was wrong (the "it's only value if it wins crowd").

Again as in Bayes experiment, if this same person found 10,000 bets with such a positive expectation, and the numbers showed the underlying theory was sound, the result would be that each outcome would be seen as having a large random element, so whether or not the next bet wins or loses isn't a large concern, it's a long term game, that's what value is about.

I believe that the value of any truth is dependent on it's use to the person who holds it. If you believe that "it's only value if it wins", to me this implies a belief in determinism (an all intelligent being would be able to say before an event starts what the outcome will be with 100% accuracy).

Accepting value, means embracing uncertainty, either in the form of our own collective ignorance or true randomness. Even if it was the case that an event could be predicted with 100% accuracy by an all-intelligent being, it wouldn't make any difference because neither you or your competitors can do this, so it becomes a race to get 'as close as possible' to the truth, this is where the value concept comes in.

Again opinion is rigid, value is mutable.

Coming back to the premise that the value of any truth is dependent on how useful it is, simply look at the most successful people in gambling / trading / investing and any other domain where value comes into play and see if they believe in value or not!

simply look at the most successful people in gambling / trading / investing and any other domain where value comes into play and see if they believe in value or not!

Survivor Bias would have a significant influence in those results.

The point about value meaning nothing without winners is, in fact, correct. If after 10,000 bets you've had no winners, has value been found? Value models are built on predictions, just as gut-feelings are built on predictions. The only difference is one of long-term averages, both of strike rate and of price.

Does 'value' really mean nothing without winners?
If I offered 20/1 on Vettel winning the F1 championship this year would anyone deny that would be a 'value' bet to take?

Also, if you toss a fair coin enough times a run of 10,000 heads in a row will surely happen eventually. :)

Betfairy, I don't understand where survivorship bias comes into play. Wouldn't this apply equally to all participants, not just those who believe in value?

If after 10,000 bets you've had no winners, has value been found?
It's reasonable to assume that in this case the numbers have not shown that the underlying theory was sound, so no. But your case is obviously an extreme one.

Let's say you spent your time backing 1.1 shots, you can expect a strike rate around 90.9%, but as you believe the true odds to be around 1.08, your expected strike rate would be around 92.6%.

If after 10,000 bets your strike rate is 91%, does this mean you were too optimistic regarding the size of your edge? Probably yes, but there is no proof, only likelihood. You could go even further and assert that you have no edge and that the 0.1% deviation from the expected strike rate is random.

Value means nothing without winners.

If there are no winners, the only logical conclusion is that there never was any value in the first place (unless the sample size is too small)

The point about value meaning nothing without winners is, in fact, correct. If after 10,000 bets you've had no winners, has value been found?

I could also turn this question around:

If someone that believes in 'backing winners' and not in value has made a highly consistent profit over a series of 10,000 bets, has this person in fact been backing value?

The better the odds are that you obtain, the higher your margin of safety.

Another one for the 'price doesn't matter crowd'.

Let's say that you know that a certain game will end in a draw (in the real world you cannot actually know this, you can only be highly confident) as it is fixed.

Does price matter. Do you care if you can back the draw at a price of 1.5 or 3? Or do you think it doesn't matter because you will win anyway so value isn't important??

An apocryphal cautionary tale:

I know a guy who collects £5 notes and £10 notes. He is willing to buy them for £20 each. My friend and I each have £50 in 6x£5 notes and 2x£10 notes. I went to swap my 2x£10 notes for for 4x£5 notes before trading them for £20 each, so that I could get £200 instead of £160 for my £50.

My friend settled for £160, because value isn't important to him.

price doesn't matter crowd

I didn't say price doesn't matter. I did say that "value" with no winners is no value at all. To formulate a price for seeking value a predictive strategy is used - a projected long-term average. If the predictions are wrong, so is the estimation of value.

Is your current betting strategy not simply to lay the outside correct scores, then hope that goals are not scored in quick succession? What does that have to do with value? Seems more like a simple trading strategy to me.

Betfairy, I don't understand where survivorship bias comes into play. Wouldn't this apply equally to all participants, not just those who believe in value?

There are so many investing strategies for stocks it would take a lifetime to list (one for every person). To simply imply that only value-seekers are long-term winners in business is naive. Take Warren Buffet. His "value" estimations may be roughly in-line with other investors, yet he invests anyway and hold long-term, helping build the business. So were all the other investors wrong not to invest? However, perhaps Survivor Bias was the wrong term to use.

What it comes down to is this. Can sports/horses be accurately predicted like the toss of a coin/roll of the dice can be? If so, why doesn't everyone win? If not, are we just betting on a flock of swans, hoping that none is actually black?

Can sports/horses be accurately predicted like the toss of a coin/roll of the dice can be? very nearly

If so, why doesn't everyone win? Because those who bother to put the effort in to do the calculations don't tell those who don't bother

In other words, your predictive skills are better than average. Hence it's a predictive game.

If you've ever watched people urinate their lives away at a sic bo table, you'll know what I mean and that's simple dice.

I just bet when the odds are wrong, it's really as simple as that.

You know they're wrong because your predictive model tells you so.

The Betfairy 29 Mar 17:30

Betfairy, that last bit wasn't aimed at you, there were others who literally said 'price doesn't matter'.

I didn't say price doesn't matter. I did say that "value" with no winners is no value at all. To formulate a price for seeking value a predictive strategy is used - a projected long-term average. If the predictions are wrong, so is the estimation of value.

I agree with that. But like I said, believing you have value is not the same as having value.

Is your current betting strategy not simply to lay the outside correct scores, then hope that goals are not scored in quick succession? What does that have to do with value? Seems more like a simple trading strategy to me.

Not exactly, I'm also happy to back outside correct scores if I think there is value there, although it is true that it is usually damaging for me when goals are scored in quick succession. The simple reason for this is that if you bet in the direction of more goals, you need to revise your bets constantly as you are in danger of being matched as the price move against you. A strategy whereby a large win due to a flurry of goals is equiprobable to a large loss would be superior to my own, but impossible to implement for me using a manual trading strategy as the number of markets I cover would not make it feasible to adjust all my bets every 20 seconds.

As it happens, I do't believe for one minute that sports are mathematically predictable, though how the averages break even is still a paradox to me.

Example1: Chelsea 7-1 Aston Villa. Play that game 100 times and Chelsea will win 100 times.

Example2: Gold Cup. Two short-priced favourites thrashed. Run it 50 times and you'd get the same result.

So how can so may predictive models be so wrong about these two results?

Example1: Chelsea 7-1 Aston Villa. Play that game 100 times and Chelsea will win 100 times.

Example2: Gold Cup. Two short-priced favourites thrashed. Run it 50 times and you'd get the same result.

I do hope you're kidding

So how can so may predictive models be so wrong about these two results?

For the same reason that you can throw three dice and it's impossible to throw the exact average.

Okay, maybe the Gold Cup was a little extreme, but the Chelsea result not so.

If Chelsea v Aston Villa were to be replayed under the exact same conditions, would you offer 1.33 on Chelsea? Or would you be happily back @ 1.02?

Given the new information, I'd lay 1.25 no problem. But I doubt I'd get matched.

What was the price before the game?

If I had this piece of additional info it might be enough to move the price from 1.33 to 1.32 or something, soccer isn't my strong point, but I wouldn't be backing 1.02

If I had this piece of additional info it might be enough to move the price from 1.33 to 1.32 or something

You're probably right. And that's why there will always be room in this betting world for a well opined "gut-feeler" to turn a profit.

FWIW There's no way in this world any of us should be arguing with someone called "Bayes" about probabilities given a previous price coupled with observational evidence.... imo :D

Replace "value" with "better than average" and I'll agree :)

Would you back or lay the 1.32 if the price shortened from 1.33

I bet with my gut for 15 years with no computer help and my gut says it's a lay.

The Betfairy 29 Mar 17:57
Replace "value" with "better than average" and I'll agree :)

We're starting to get bogged down in semantics ;)

Interesting about Bayes. I consider sports probabilities much more akin to Bayesian probability than frequentist probability.

"I believe my model to be accurate, and I won't be seeing any of those black swans for some time"

Bayes did my head at uni :(

We're starting to get bogged down in semantics

Value is a dirty word in these parts :)

Replace "value" with "better than average" and I'll agree

it is proven that BFSP = average probability

so beat BFSP is the answer.

^^^ I still find that quite remarkable.

Lex: so beat BFSP is the answer

Or, for commission management reasons, on average beat bfsp with yr bets and green up laying bfsp at the off.

But I wd hazard it's a lot harder than it sounds.

Everyone looks for value, the ones that say they don't do it without even realising it. E.g. someone backs Chelsea at 1.3 because they think they are a 'cert' are effectively saying that they think the true odds are 1.01 or something near.

The people who don't believe in value or don't acknowledge that they are subconsciously looking for it probably bet varied amounts based on their confidence on the bet. The more confident they are the more they bet and the more confident they are the closer they are to believing a bet is nearer 1.01. This is in effect Kelly Staking!!

That said you don't necessarily have to have value in every bet and you realistically won't. By staking more based on confidence you are more likely to be in profit than someone betting at level stakes (if you really are good) but at the same time you are probably suseptible to overstaking and having false confidence in your bets mainly becasue you have no records or real structure to your betting.

Also, someone who bets with value may well lose after 10 bets and then even over 100 but after this it becomes very unlikely that you won't be in profit, unless you never really had real value in the first place!

Betfairy: I still find that quite remarkable.


Have a look at wisdom of crowds.

I really can't understand what you are saying about the gold cup and the chelsea game.

You truly are a black swan yourself, IMO.

Good Value thead with the usual contributors.

I like reading value threads, I'm always amazed by some of the ludicrous opinions, even some of those that claim to understand the simple concept of value don't get it, i.e. Schalke.

I agree, it's the people that seem to know what value is but then deny it has any importance.  A bit like scientists that believe in god.

Not another non-religion debate, Cat!!   That thread was even more ludicrous than this one!!

I get it in wherever I can LordB .....

How do you pull up threads from 2 nearly years ago ?

I thought this was going to be good suggestions on restaurants worth visiting-poor title in my opinion!!!!!

How do you pull up threads from 2 nearly years ago ?


kenilworth, i've got half a dozen old threads saved ('add to favourites' button at top of thread)....that's one way, if it had been saved

Mar 1, 2012 -- 5:01AM, duncan idaho wrote:

How do you pull up threads from 2 nearly years ago ? kenilworth, i've got half a dozen old threads saved ('add to favourites' button at top of thread)....that's one way, if it had been saved

do u keep threads as favourite for 2 years?

On the Beta Forum, you can sort threads by Most Replies, so the ones with 500+ posts appear on the 1st page of the forum.

Or you can use advance search on google and enter  http://community.betfair.com in the 'site or domain' box

All this of course is still rubbish compared to the old search archives we had before the 'upgrade' in 2010

do u keep threads as favourite for 2 years?


sadly, yes, as they tend to be long ones which i never find time to wade thru cos my eyes sting on here by the end of the day

zipper
27 Mar 10 12:32 Joined: 06 Jul 02 | Topic/replies: 5,766 | Blogger: zipper's blog
The only value is the Winner .. price dont matter ... trust me .



WHAT AN IDIOT YOU ARE ZIPPER .

and the other moron who mentions  6billion to one chance and it loses - no value at all .


after 1 event you cant possibly judge it

PATHETIC

Given the tendency of edges to erode over time, I'd guess that bayes's earlier 2.1 shots were better than that and maybe some at the end not so good.

Another question is the degree to which most bet selection processes pick out the same factors, which they weight more heavily than the market, time and again. My feeling is that people are operating an algorithm in their heads even when they are unaware of this.

Most factors and their weights revise v. quickly in horse racing i.e. a particular race was strong and will feature a lot of winners, the going was different than it was advertised, the draw is not the way it was first thought etc. Find something that leads to the general market being mispriced more quickly than general opinion does and you will be on yr way to striking value bets.